Universal Envelope

Purpose of the Universal Envelope

All ATP data is expressed using a single canonical structure called the Universal Envelope.

This ensures that:

• Identity events, claims, contracts, reviews, and attestations

• Can all be stored, indexed, verified, and referenced using identical mechanisms

The envelope removes the need for special-case handling of different event types.

---

Envelope Structure

{
  "id": "<sha256_hash>",
  "kind": <integer>,
  "ai_id": "<ai_id>",
  "created_at": <timestamp>,
  "tags": [[...]],
  "content": <any_json>
}

Each field has strict semantics:

• id A SHA-256 hash of the canonically serialized event. Any modification produces a new id.

• kind Integer identifier defining the semantic meaning of the event.

• ai_id The cryptographic identity that authored and signed the event.

• created_at Unix timestamp indicating when the event was created.

• tags Structured metadata used for indexing, filtering, and relationship modeling.

• content Arbitrary JSON payload defined by the event kind.

---

Canonical Serialization

Before hashing, the event must be serialized using strict canonical rules:

• UTF-8 encoding

• Alphabetically sorted keys

• No whitespace

• Unicode NFC normalization

This eliminates implementation variance across:

• Programming languages

• Platforms

• Storage layers

Deterministic serialization is critical: verification must be universal.